Jump to content
Toggle menu
Toggle preferences menu
Toggle personal menu
Not logged in
Your IP address will be publicly visible if you make any edits.

GLM 5.3 Flash on 4x 8 GB/Confirmation 5

From CMP Wiki
Revision as of 08:14, 30 October 2026 by ImportBot (talk | contribs) (Import: Confirmation of GLM 5.3 Flash on 4x 8 GB: 1a2b3c0967's 1M-context PP4 run (credited import); sources: @1a2b3c0967)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Confirmation
ConfirmsGLM 5.3 Flash on 4x 8 GB
ResultReproduced
By@1a2b3c0967
Date2026-10-04
Cards8 GB
NumbersRelease 1.7, PP4, concurrency 3, 1M context, KV pool about 3M tokens: over 75 tok/s on prose.

Dropped --max-num-seqs from 8 to 3 to grow the KV pool. Source: @1a2b3c0967, 2026-10-04